Glioblastoma multiforme (GBM) is the most common primary malignant tumor of the adult central nervous system (CNS). The median survival after surgical intervention alone is approximately six months and the addition of radio-/chemotherapy can extend this time up to twelve months. Failed therapy is most often associated with local recurrence in the proximity of the original tumor. The limited clinical benefits of currently available treatments warrant the consideration of novel approaches for GBM. In this regard, the recent application of replicative viral systems (CRAd) represents such a novel therapeutic approach. As an adenovirus (Ad)-based cancer gene therapy approach, CRAd oncolytic potency is dependent on vector-mediated tumor infection. However, studies have demonstrated that malignant glioma are especially resistant to adenoviral vector infection, due to a deficiency of its primary receptor, the coxsackie and adenovirus receptor (CAR) on brain tumor cells. In addition, the lack of well defined tumor-selective promoter (TSP) elements that demonstrate selective transcriptional activity in glioma has hampered the construction of CRAds with desired replicative specificity. To this end, we have developed novel vectors that are genetically modified to encode highly specific targeting ligands capable of mediating enhanced vector transduction of CAR-deficient cells by virtue of specific targeting of an alternate cell-surface receptor. For GBM, our specific targeting of glioma cells will be based on the selective expression of the IL-13a2 receptor on malignant brain tumors. Application of our transductional targeting vector technology will involve the incorporation of a mutant form of hIL-13, IL13.E13K (muIL13) into the Ad vector capsid to function as a glioma-specific targeting ligand. This form of hIL13 has shown negligible binding to normal cells and even higher affinity to glioma-specific IL-13a2 when compared to hIL13. Furthermore, we have tested a variety of TSPs, and have identified the survivin (S) promoter as the optimal transcriptional control element for control of E1a, a gene essential for CRAd replication. Importantly, survivin expression in gliomas is associated with poor prognosis, increased rates of recurrence, and resistance to chemo- and radiotherapy. Based on the above considerations, it is clear that employment of the combination of IL-13a2 targeting and survivin controlled replication will result in the derivation of a CRAd agent of utility for GBM. We hypothesize that an Ad vector targeted to IL-13a2, when rendered selectively replicative via the surviving promoter, will demonstrate transductional and replicative specificity required for human clinical trials and thereby allow full realization of the potential benefits of a CRAd approach for GBM. Public Health Relevance: Glioblastoma multiforme (GBM) is the most common primary malignant tumor of the adult central nervous system (CNS). The median survival after surgical intervention alone is approximately six months and the addition of radio-/chemotherapy can extend this time up to twelve months. As a result of recent identification of IL13a2 receptors on the surface of glioma cells, we now propose to develop a conditionally replicative adenoviral vector (CRAd) which utilizes a glioma specific promoter-survivin-and binds selectively to glioma cells that express IL13a2.